Prediction for cloud computing.
I believe private cloud is dead - you don't need it. If you have a virtualised infrastructure already, that's good enough, you just need decent deployment tools and even the standard tools supplied by VMWare, Microsoft and others tend to be good enough. However, one might want to consider using containers (LXC or Docker) but beyond that, private clouds don't give you anything above what virtualisation technologies do.
Public clouds have been only good for a handful of 'consumer' applications. You're only now seeing the likes of IBM, SAP and Oracle slowly moving their apps into a cloud model - so when they mature in a few years, that would be the time to start considering them. Having said that, as others claim on here, public cloud is not for everyone due to security concerns/laws.
Hybrid clouds might work better, but not in their current guise. I think once what they offer can be componentised (not an entire application, but a subset) into offering things like Docker in a seamless and dynamically interchangeable model - whereby application subsets are loaded and offloaded, into/out of the private side of a hybrid cloud, as the business needs change - then you'll be in a real position to consider a move to cloud computing. I'm predicting a capability in a hybrid cloud environment, whereby application subsets (based on something like LXC or Docker), come in and out of one's private cloud (from a public cloud - probably supplied by an apps vendor) to deliver a particular function or application service. This movement of apps/subsets creates a hybrid cloud (because you have a mixture of app types - in-house developed/running and those coming in from a public cloud) and breaks things down to beyond an application; actually using subsets delivered by something like LXC or Docker.
Until then, it's all marketing and a repeat of what one already has at the moment = virtualisation.
Biting the hand that feeds IT
